The incidence of epilepsy rises rapidly after age 60, and many elderly are being treated with phenytoin (PHT), carbamazepine (CBZ) and valproic acid (VPA). These older antiepileptic drugs (AEDs) have many shortcomings including complex metabolism by the cytochrome P450 and glucuronidation enzyme systems and have been shown to be inducers and inhibitors in these systems. This makes them prone to many drug interactions involving both clearance and protein binding. This is a multifaceted issue; AED efficacy and toxicity may be altered by co-medications, and AEDs can affect the efficacy and toxicity of co-medications. Because an elderly patient uses an average of 6 medications, the risk of medication interactions in this age group with these older AEDs is very high. Three newer AEDs, lamotrigine (LTG), topiramate (TPM) and levetiracetam (LEV) appear to have more favorable drug-drug interaction profiles; all have low protein binding and fewer or no metabolic interactions. However, these newer drugs have not been studied sufficiently in the eldedy and more detailed information regarding the pharmacokinetio and pharmacodynamic properties of these is needed. The difficulty in obtaining blood samples from this population makes inclusion in standard pharrnacokinetic studies difficult. This project will use nonlinear mixed effects model (NONMEM) that employs both pharmacokinetics and statistics will be used to determine pharrnacokinetic parameters of these three new drugs. This powerful method allows the use of routinely collected data to be used and avoids the risks and expense encountered in intensive pharmacokinetic studies. With this method, not only can the drug clearance be determined for a population, but it can also be determined for an individual. Factors (age, race, gender, smoking, etc.) that affect drug clearance can also be determined. In addition, the relationship between serum drug concentration and seizure type will be determined for LTG, TPM, and LEV. We will have access to approximately 450 persons >65 years of age receiving LTG, 420 receiving TPM and 337 receiving LEV from several active epilepsy practices in 3 cities (Minneapolis, Miami, Atlanta) and data from more than 1500 younger adults on each of these AEDs. In addition we will use our tools to analyze data from the ongoing perspective VA cooperative study of LTG projected to enroll 240 subjects receiving LTG as initial treatment. The VA data will determine the relationship between drug concentrations and adverse events and seizure frequency for LTG providing both pharmacokinetic and pharmacodynamic information in the elderly. This project along with Projects 1 and 2 will provide pharmacokinetic data and identify and quantitate the factors that influence the pharmacokinetics of LTG, LEV, and TPM. This information can be used to guide dosing requirements needed to obtain target serum concentrations in the elderly to achieve seizure control and avoid drug toxicity.